Surveillance camera systems are often deployed to collect image data within or around premises. Examples of premises include governmental buildings, office buildings, retail establishments, and single and multi-unit residences. The cameras are typically installed to monitor and detect individuals and/or activities at different locations in and around the premises.
A successful installation of surveillance camera systems requires careful consideration of several factors. The designers/installers select the locations at which to install the cameras, select the type of camera that is best suited for each location, and then position the cameras' fields of view to capture scenes at each location. For example, point of sale areas might require one or more ceiling mounted, dome style cameras to capture transaction-related activities within the locations. For monitoring large open areas such as shopping malls, open-floor plan offices, and parking lots, either panoramic view (e.g. “fish eye”) cameras or pan-tilt-zoom (PTZ) cameras are often utilized because of each camera's ability to provide wider fields of view and to scan the areas, respectively. Designers/installers might also position the fields of view of different surveillance cameras to overlap, and also position the field of view of one camera to include another surveillance camera. These actions provide different views or perspectives of the same scene and the ability to capture attempts at tampering with the surveillance cameras.
Analytics systems are often part of surveillance camera systems. At a basic level, the analytics systems provide the ability to detect and track individuals and objects within the image data of the monitored scenes, Other capabilities include the ability to determine motion of objects relative to visual cues superimposed upon the image data and to search for specific behaviors of interest within the image data. The visual cues are often placed near fixed objects in the background scene of the image data to infer motion of objects relative to the visual cues. In one example, virtual tripwire visual cues can be located near entryways within the scene to detect entry or exit of individuals through the entryways and to provide a count of the individuals passing through the entryway over a specific time period. These analytics systems can provide both real-time analysis of live image data and forensic analysis of previously recorded image data.
A common problem when installing surveillance cameras is the potential for deadzones, which are areas in the premises that are not included within any of the fields of view of any of the cameras. These deadzones might result from improper positioning of the cameras and/or improper selection of the type of cameras installed. On the other hand, the existence of deadzones might simply be the consequence of a limited budget and the fact that deploying enough cameras to achieve total coverage is not feasible. At the same time, deadzones can result when there are obstructions present in the scenes of the cameras. Obstructions within the scenes such as pylons, bookcases, cubicle walls, file cabinets, and stairwells can cause areas behind the obstructions to be hidden from view. These deadzones remain unless additional cameras can be added that provide different views of the obstructed areas.